But of course the biggest risk of all in false-positives is that they open up the door to more invasive tests such as amniocentesis, which carry their own substantial risks. Ultimately, more healthy babies could be lost through amniocentesis than Down’s babies be detected.
Considering its extremely poor track record of accuracy, ultrasound may not even be useful very late in pregnancy to help confirm the position of the baby. The only rationale for scanning most uneventful pregnancies is to satisfy our curiosity, to try to get a little closer to the mystery of life.
FOETAL MONITORING DURING LABOUR
When labour commences and you go to hospital to have your baby, chances are that doctors will want to strap you into an electronic foetal monitor to routinely monitor the foetal heart rate. However reassuring, this exercise appears to make no difference to the health of either you or your baby, according to British research. Interpretation of the read-outs varies widely among practitioners. What foetal monitoring does do, however, is increase the likelihood of obstetrical intervention. Women given EFM have more than a 5 per cent higher likelihood of a forceps-type delivery and a greater chance of having a caesarean than those who employ a Doppler (handheld) ultrasound device, according to British research.43
The New England Journal of Medicine reluctantly concluded, after examining seven major studies, that this form of ultrasound provides no benefits to newborns, even premature ones. In reviewing the data, the Journal accepted that the study was the final proof that foetal monitoring is ineffective in decreasing the chances of a stillbirth, a low Apgar score or neurological problems in high-risk infants. It only increases the chances of a woman having a caesarean section.44 This conclusion was reached following a study, carried out in several medical centres in the state of Washington, which tested the widely-held view that high-risk babies who were electronically monitored died less frequently and had better outcomes than low-risk babies monitored by simple auscultation (trumpet stethoscopes) or other sonic aids. The study, which looked at premature infants in several hospitals, found that those babies monitored had no better chance of being born live than those monitored by ordinary Doppler auscultation. The final death-knell was sounded when a major California study found that the test’s false-positive level – reporting a problem where there is none – is an alarming 99.8 per cent, resulting in thousands of unnecessary caesareans.45
Even the former head of the Oxford Perinatal Unit, Iain Chalmers, has gone on record to say that the major, properly conducted studies show that the mortality rate among technologically monitored babies was higher than that among controls.46
‘A review of this evidence was first published eight years ago,’ wrote Chalmers,47 ‘and the lack of evidence to support the use of this widely adopted form of obstetric technology has been reiterated at intervals since then.48 For obvious reasons, it is the kind of evidence that some obstetricians would rather overlook.’
Unless you are suspected of having twins, you might be wise to avoid knee-jerk ultrasound testing, particularly before your 20th week of pregnancy when the baby is still forming.
AFP TESTS
Most of the rest of the new prenatal tests are designed to detect Down’s syndrome, and new ones are devised as fast as some of the older ones seem to get discredited – this despite the fact that none of these screening tests seems to be making much difference. Despite 30,000 amniocenteses and 3,000 chorionic villus tests performed each year in the UK, less than 20 per cent of Down’s syndrome babies are detected. This may have something to do with the fact that 70 per cent are born to younger mothers who don’t have the tests – a fact that tends to pour cold water on the idea that Down’s syndrome is solely a result of the ‘tired eggs’ of relatively elderly mothers.
In fact, despite medicine’s attempts to protect mothers from having Down’s babies, the incidence of the condition is going up. This could either be because the tests – usually amniocentesis or alpha-fetoprotein – are not detecting the condition, or because parents are choosing not to abort the babies diagnosed as suffering from the condition.
The most substantial risks you face of deformed or retarded children may result from the diagnostic tests themselves.
AFP Test
Before you go in for amnio or CVS, you’re likely to have, as I did, an AFP test or a ‘triple test’, developed by the University of Leeds to replace advancing maternal age as the only risk factor for giving birth to a Down’s syndrome baby.
AFP stands for alpha-fetoprotein (AFP), a protein produced by the foetal liver. If the baby has a spinal abnormality such as spina bifida or anencephaly, where the full brain is not developed, larger amounts of AFP than normal will be present in the amniotic fluid and the mother’s blood. Low levels of AFP are considered a possible indicator of a baby with Down’s syndrome. The test is given between a woman’s 16th and 20th week of pregnancy.
There is no doubt that the batting average on the alpha-fetoprotein tests is appalling. After my own experience, I heard of at least three friends or acquaintances with false-positive AFP readings. Doctors accept there is a 3–4 per cent error rate of abnormally high readings on first screening, according to writer Helen Klein Ross. ‘This means that of every 2,000 women tested,’ she says, ‘100 will have an abnormal reading, but only 1 or 2 will be carrying a foetus with this congenital defect.’49
Even this estimate of inaccuracy could be conservative. One study done in 1982 estimated a failure rate of 20 per cent.50 As with many biological processes, AFP readings can change from day to day or be falsely elevated, depending on many factors: multiple pregnancies, viral hepatitis, smoking, a threatened miscarriage or even carrying a boy all can produce a false-positive, and being overweight, having insulin-dependent diabetes or just being wrong about your date of conception can produce an artificially low score51 (some 25 per cent of all inaccurately high AFP results are due to inaccuracies about dates or a multiple pregnancy52). Certain races, notably Asian and black women, also seem to produce abnormally high AFP scores.53 According to the late Dr Robert Mendelsohn, one of the first to call attention to the problems of this test in his American newsletter The People’s Doctor, the test has false-negatives too, as evidenced in an article in The Lancet concerning two babies born with spinal defects whose mothers nevertheless had normal AFP readings.54
According to Helen Ross, the AFP test ‘misses about 40 per cent of spina bifida cases, 10 per cent of anencephaly cases, and 80 per cent of foetuses with Down’s syndrome. All of which makes a negative result by no means reassuring.’55
Twins or a miscalculation about the date of conception are two common ways that test results are thrown off. In my case, we were sure about the dates, but my daughter Caitlin turned out to be a full 10-month pregnancy, born 28 days past her estimated due date (first babies who are not induced are very often late, and my obstetrician doesn’t induce labour if there is no evidence that anything is awry). As a slow grower, she probably deviated sufficiently from the norm to show up as ‘abnormal’.
In other words, mostly what this test produces is a good deal of needless anxiety, which can only be dispelled by subjecting your baby to amniocentesis or ultrasound, two procedures with their own potential risks. Indeed, for anyone younger than about 39, the risk of losing